Traffic Simul

Inspiration

The daily struggle with traffic, making some of us drive over an hour to school, gave inspiration for our project. Many of us face the frustration of waking up early to commute long distances, only to be further delayed by heavy traffic. We believed there had to be a more efficient way to manage traffic flow, leading us to develop Traffic Simul.

What it does

Traffic Simul is a Unity and Python demonstration of an AI-powered traffic light system. It showcases how intelligent traffic management can potentially reduce congestion and improve commute times for everyone.

How we built it

Our system integrates two AI components: (1) A YOLOv8 model for real-time vehicle detection at intersections. (2) A custom-trained Neural Network that optimizes traffic light timing based on vehicle count data. We then implemented these AI systems with a Unity environment to create a visual demonstration of how our solution would function in real-world scenarios.

Challenges we ran into

Our journey was not without its challenges. One teammate faced difficulties installing Unity, a process that took two hours to resolve. Overcoming socket size limitations and addressing packet loss to establish a stable Python-Unity connection proved time-consuming as well. Additionally, training a neural network with limited compute resources and within a tight timeframe forced us to maximize every minute we had.

Accomplishments that we're proud of

Despite our initial doubts about completing the project within our 8-hour coding window, we were eventually successful in integrating multiple complex technologies (YOLOv8, Custom Neural Networks built from the ground up, and Unity) to create a functional demo.

What we learned

We learned a lot about, Unity simulation development; AI model implementation, integration, and development; and real-time efficient data processing and system optimization.

What's next for Traffic Simul

To further develop Traffic Simul we would like to simulate more complex and diverse traffic, like on highways and complex intersections, and develop a complex AI structure to manage all traffic lights in a city.